This study constitutes the first report of TB treatment outcomes in a large HIV-TB co-infection cohort in India. Globally and in India, TB is one of the most common opportunistic infections affecting people with HIV. This assumes importance in a country like India which has 2.7 million HIV infections and 23% of the world’s incident TB cases. HIV infection is often cited as an important reason for failure to control TB, and for causing a resurgence in TB worldwide. While this is true, our results suggest that implementation of program guidelines in a coordinated manner can result in good treatment outcomes among those co-infected with TB and HIV.
In this study, pulmonary involvement occurred in about 73% of all HIV infected patients with TB. This is in line with previous reports from India . Smaller studies from Thailand have reported a lower proportion of pulmonary TB (60%)  while a report from Kenya  stated 87% co-infected patients had pulmonary TB. While overall treatment success rates in our co-infected cohort were close to 75%, they were significantly better in the group that had been already initiated on ART compared to those who were not on ART. Similarly, mortality rates were twice as high in the group that had not started ART at the time of TB treatment. Recent studies from different countries have repeatedly shown that early initiation of ART among co-infected individuals even during the intensive phase of anti-tuberculous therapy can decrease mortality among those with HIV/TB co-infection [15–18]. While from a biological point of view, patients on ART have lower viral replication and hence better host responses, from a program point of view, these patients are regularly clinically followed at the ART centers where adherence is repeatedly emphasized, possibly contributing to better outcomes in this group. This finding in our Indian cohort lends local contextual support to the recently introduced (Nov 2011) Indian program guideline of initiating ART in all HIV-TB co-infected individuals, as soon as possible in the intensive phase of anti-tuberculous treatment, regardless of CD4 counts . A recent study from the same province in India has shown that only 21% of a cohort of co-infected patients was on ART at the time of TB diagnosis . The same report suggests that despite operational weaknesses in the program, ART could be extended to all HIV-infected TB patients irrespective of CD4 count with relatively little additional burden on the national HIV control program .
Treatment success proportions in the co-infected cohort were similar to those achieved within the routine TB control program in the province. There was an expectedly significantly higher death rate among the co-infected patients (16% versus 7%) in the province. Globally, TB-associated mortality in co-infected patients is three times higher than mortality among TB only patients. There are a number of possible explanations that have been proposed for the increased mortality among co-infected patients. The location and extent of TB is influenced by the degree of immunosuppression, often increasing the difficulty of diagnosis and hence delaying treatment initiation, resulting in higher mortality . Immunological studies have also shown the host responses to M. tuberculosis enhance HIV replication [22, 23], thus accelerating the natural progression of HIV and further depressing cellular immunity. Decreased gut absorption of anti-tuberculous drugs has been suggested by some reports , leading to impaired treatment outcomes including death. Non-initiation or delayed initiation of ART in this cohort has also contributed to higher mortality . Our analysis also indicated that rates of default and failure were significantly higher among TB only patients than in the TB HIV co-infected cohort. This is possibly because of a significant proportion of re-treatment patients entering the TB program (20% of all registered patients) compared to the co-infected cohort. In addition it is likely that the emphasis on multidisciplinary care and adherence support to patients within the ART program may play a role in contributing towards the lower rates of default and failure among the HIV/TB co-infected group, particularly among those on ART.
This study is limited by the non-inclusion of the patients taking ART privately. Only co-infected cases within the ART program in the province are considered in this study, hence the number of and outcomes for such cases not registered in the ART program cannot be ascertained. However this excluded group is likely to be small. The same is also the case for patients receiving TB treatment outside the program. The exact time duration between initiation of ART and ATT cannot be identified reliably from this dataset. Another limitation of the data set is the absence of information on sites of extra pulmonary infection. Bias is a problem as co-infected patients are included in the main TB program registers (co-infected patients are not removed from the registers and reported separately), though in the summary statistical reports from the TB program used in this study, they are not identifiable as such. The strength of this study lies in the large size of the co-infected cohort, and the results are likely to represent the majority of those accessing HIV or TB care in the country.